A classification method of light-variable curves based on support vector machine

Abstract

The invention discloses a light change curve classification method based on a support vector machine, which comprises the following steps: Step 10) collecting light change curve data and light changecurve types; Step 20) preprocessing the collected light change curve data; Step 30) constructing a data set to be analyzed according to the preprocessed light change curve data, and dividing the dataset into a training set and a test set; Step 40) setting kernel function and punishment coefficient of the support vector machine, training initial classification model by using the constructed training set to obtain the trained classification model, testing the correct rate of the trained classification model by using the constructed test set, and taking the trained classification model whose correct rate reaches the set threshold value as the final classification model; Step 50) classifying the light curve by using the final classification model. The method has strong anti-interference ability to the noise signal, more accurate classification, and can detect the abnormal light curve type at the same time.

Description

technical field [0001] The invention relates to the field of astronomical data observation and analysis, in particular to a method for classifying light curves based on a support vector machine. Background technique [0002] For a long time, astronomical research approaches mainly include spectroscopy and imaging in the optical band. The traditional research methods are manual selection of several targets, tracking and shooting, data processing and analysis. This research method is inefficient, and the astronomical community has been in a state of data scarcity for a long time. Driven by emerging technologies such as information and computing technology, the field of astronomical research has shifted from the traditional mode of observation with fewer targets and manual data processing to a data-intensive era. A large number of sky survey projects have provided a large amount of data for astronomical research, such as ROTSE, ASAS, SuperWAS, MACHO, OGLE, SDSS, LAMOST, and K...

AI Technical Summary

Problems solved by technology

Kepler uses a polynomial to fit the light curve, and gives the light curve type according to the width and depth of the main minimum and sub-minimum of the fitted curve; both ASAS and ROTES use Fourier transform to extract the frequency characteristics of the light curve data , classify according to the proportional relationship between DC, 2nd order and 4th order frequency components in the obtained frequency values. Due to the use of less feature quantities, it is easy to be affected by data fluctuations caused by instrument test errors and weather reasons. For abnormal astronomical phenomena Abnormal light curve lacks recognition ability

Images